Hi Phillip, Why would you like to use Apache Stanbol? What would be the unique selling point?
The only thing I could see is custom entity linking. For those who are interesting this is the way I do it * Index your Entities in Apache Solr - so use Solr to replace the Stanbol Entityhub * Use the Apache Solr Tagger Handler https://lucene.apache.org/solr/guide/8_5/the-tagger-handler.html for linking * Feel free to use the SolrJ Client https://github.com/redlink-gmbh/solrj-text-tagger OK it is plain entity linking - meaning that every single token is a candidate for linking. Stanbol could do things like only linking Nouns or Named Entities ... but for most use cases plain entity linking is sufficient. If more advanced linking modes are needed just add some NLP processing and check if Tags created by the Solr Tagger overlap with Words classified as Nouns/Numbers or Named Entity Annotations. BTW the Solr Tagger is version 2.0 of the Code used by the Apache Stanbol FST linking engine. It was contributed to Apache Solr with version 7.4. If you are stuck with an older Solr Version have a look at https://github.com/OpenSextant/SolrTextTagger best Rupert On Fri, 3 Jul 2020 at 07:21, Phillip Rhodes <motley.crue....@gmail.com> wrote: > > N/M, I see Stanbol got moved to the attic already. Shame. > > > Phil -- | Rupert Westenthaler rupert.westentha...@gmail.com | Bodenlehenstraße 11 ++43-699-11108907 | A-5500 Bischofshofen | REDLINK.CO .......................................................................... | http://redlink.co/
